Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Myelodysplastic syndromes (MDS) and myeloproliferative neoplasms (MPN) are a heterogeneous group of clonal stem cell disorders that result in ineffective hematopoiesis and an increased risk of developing acute myeloid leukemia (AML). Whereas in MDS ineffective hematopoiesis results in progressive cytopenias, in MPN ineffective hematopoiesis leads to excessive proliferation frequently characterized by increased peripheral blood counts. MDS and MPN are rare in children, and the only malignancy with overlapping features of both MDS and MPN in childhood is referred to as juvenile myelomonocytic leukemia (JMML). Chronic myeloid leukemia (CML), characterized by the Philadelphia chromosome ( BCR/ABL- positive), is more common in adults and seen infrequently in children. However, other forms of MPN including polycythemia vera (PV), essential thrombocythemia (ET), primary myelofibrosis, chronic neutrophilic leukemia, chronic eosinophilic leukemia, chronic basophilic leukemia, chronic myelomonocytic leukemia, systemic mastocytosis (SM), and stem cell leukemia–lymphoma syndrome) are exceedingly rare in children. Additional information regarding disorders that are predominantly found in adults are referred to in Chapter 70, Chapter 71, Chapter 72, Chapter 73, Chapter 74, Chapter 75 . In contrast, two MPNs are uniquely pediatric: JMML and Down syndrome–associated transient abnormal myelopoiesis (TAM) will be the main focus of this chapter.
The MDS are a heterogeneous group of clonal disorders characterized by ineffective hematopoiesis, impaired maturation of hematopoietic cells, progressive cytopenias, and dysplastic changes in the bone marrow (BM).
MDS is a common malignancy of adults, with an incidence of 50 cases per million in people older than the age of 60 years. In contrast, MDS accounts for only 3% to 7% of all hematologic malignancies in children, with an unknown true incidence, owing in part to the inclusion of children with Down syndrome in some estimates. Several population-based studies have been performed with reported incidences of 4.0 cases per million in Denmark, 3.1 per million in British Columbia (Canada), and 1.35 per million in the United Kingdom. The median age of presentation is 6.8 years with an equal sex distribution. However, in a study of children with advanced or high-risk MDS, the median age of presentation was older, at 10.7 years with a 2:1 male-to-female ratio.
As in adults, MDS in children can be considered as primary (de novo) or secondary. In adults, two predominant patterns of primary, de novo MDS have been observed. In the first of these patterns, the disease is indolent in nature and is characterized by prolonged survival, little accumulated genetic damage, and a low probability of progression to AML. This group of diseases is best exemplified by the 5q− syndrome, an entity not seen in children. Far more common in adults is a disease characterized by the accumulation of genetic changes, progression to BM failure, and a high probability of developing AML. This form of the disease is characterized as a mutator phenotype. The primary MDS seen in children appears to share this mutator phenotype.
As in adults, secondary MDS in children can also arise as sequelae from exposure to chemotherapy and radiation, and from genetic conditions including Down syndrome, paroxysmal nocturnal hemoglobinuria (PNH), neurofibromatosis type 1 (NF1), Bloom syndrome, and Li-Fraumeni syndrome. However, MDS in children may often result from inherited constitutional BM failure syndromes including Fanconi anemia (FA), severe congenital neutropenia, Shwachman-Diamond syndrome, congenital amegakaryocytic thrombocytopenia, dyskeratosis congenital, Diamond-Blackfan anemia, and MonoMAC syndrome.
Ongoing studies are now defining the molecular pathogenesis and interrelationships among MDS, MPN, and AML. Accumulating data suggest that aberrant signal transduction resulting from acquired somatic mutations encoding proteins, leading to hyperactivation of the Ras pathway, may stimulate proliferation without concomitant differentiation. This has been clearly demonstrated in CML ( BCR-ABL ). A number of other putative pathogenetic mutations have been identified in other myeloproliferative disorders, including JAK 2 V617F and CALR mutations in PV, ET, and primary myelofibrosis ; KIT D816V mutations in SM ; FIPL1-PDGFRA in chronic eosinophilic leukemia-SM ; ZNF198-FG4FR1 mutations in stem cell leukemia–lymphoma syndrome ; RAS/NF1/PTPN11 mutations in JMML ( Fig. 65.1 ) ; and GATA1 mutations in TAM.
AML is the result of cooperating mutations in genes that confer a proliferative and survival advantage (e.g., activating mutations in receptor tyrosine kinases [FLT3, c-kit]), and genes that impair differentiation and apoptosis (e.g., loss-of-function mutations in transcription factors [CBF, AML/ETO]) ( Fig. 65.2 ). This multistep model for the pathogenesis of AML is supported by murine models, the analysis of leukemia in twins, and the analysis of patients with familial platelet disorder with a propensity to develop AML (FDP/AML syndrome).
The patterns of genetic and karyotypic abnormalities in children with MDS are typically distinct from that of adults. Although a few karyotypic abnormalities are shared (e.g., 7/7q−), many that are common in adults are only rarely found in children (e.g., 5q−). Importantly, mutations in mRNA spicing genes are only rarely found in children with de novo and secondary MDS. Mutations in epigenetic genes that control DNA methylation and histone function are also only rarely identified in children, making their role in MDS uncertain. For example, although mutations in TET2 are identified in 20% to 25% of adults with MDS, only 1 out of 19 children with refractory cytopenia of childhood (RCC) had a mutation in TET2 . Instead, the genetic abnormalities in children with MDS are most often those associated with inherited BM failure syndromes and other genetic disorders ( FANC member genes, DKC , TERT , TREC , WAS , GATA-2 , SBDS , etc.).
Finally, three exceedingly rare familial forms of MDS/AML are associated with mutations in RUNX1/AML1 (familial platelet disorder with a predisposition to AML), CEBPα (familial AML), and GATA-2 . Most recently, germline mutations in the genes SAMD9 and SAMD9L have been shown to lead to Mirage and ataxia–pancytopenia syndrome, respectively. Patients with these disorders are at a higher risk of developing MDS compared to the general population and are frequently present with monosomy 7.
Until recently, MDS in children was poorly defined, characterized, classified, and reported. In fact, MDS was not included in the International Classification of Childhood Cancer until 2005. Also contributing to this lack of information was the use of classification and prognostic systems designed for adults that have had limited applicability to children. A number of classification systems for children and adults have now been proposed. Taken together, it may be useful to think about childhood MDS as primary or secondary in nature, with secondary MDS arising either from a known inherited BM failure syndrome, prior acquired aplastic anemia, or as a complication from prior chemotherapy or radiation therapy. A diagnosis of primary MDS would then apply to all other cases.
Historically, one of the most commonly used classification systems was the French-American-British (FAB) system, originally proposed in 1982. This classification system recognized five forms of MDS in adults: refractory anemia (RA), RA with ringed sideroblasts (RARS), RA with excess of blasts (RAEB), RAEB in transformation (RAEB-T), and CMML. While RAEB and RAEB-T were commonly reported in children, RA and RARS were thought to be rare in children. However, a population-based study in the United Kingdom showed 25% of childhood MDS cases to be RA or RARS, suggesting inaccurate diagnosis or reporting of these subtypes in other pediatric studies. CMML has only rarely been reported in children.
Additional subtypes of MDS are now recognized that do not fit well into the FAB system, including hypoplastic MDS, therapy-related MDS, refractory cytopenias with trilineage dysplasia, MDS associated with myelofibrosis, and MDS associated with inherited disorders (congenital neutropenias, Shwachman-Diamond syndrome, FA), Down syndrome, NF1, and mitochondrial cytopathies. Therefore, the World Health Organization (WHO) proposed changes to the FAB criteria to account for many of these subtypes. Importantly, whereas adults commonly present with RA without cytopenias in the myeloid or platelet lineages, this is very rare in children since they more often have cytopenias in more than one cell line. Therefore, children with low-grade MDS are classified as having refractory cytopenia as opposed to RA. Thus in 2008 the WHO classification of pediatric MDS included the provisional category of RCC based on persistent cytopenia with <5% blasts in the BM and <2% blasts in the peripheral blood. Under this classification, it is recommended that children with refractory cytopenia with multilineage dysplasia (RCMD) be classified as RCC until it is clarified whether the number of lineages involved is an important prognostic discriminator in childhood MDS. Under the 2016 revision of the WHO classification of myeloid and neoplasms and acute leukemia, RCC remains a provisional entry.
Signs and symptoms of MDS are nonspecific and are usually attributable to pancytopenia (fever, infections, pallor, fatigue, bruising, and petechiae). Lymphadenopathy, hepatomegaly, and splenomegaly are uncommon presenting signs in children with MDS.
Commonly accepted minimal diagnostic criteria for pediatric MDS include the absence of common de novo AML karyotypic abnormalities and at least two of the following: (1) sustained, unexplained anemia; neutropenia or thrombocytopenia; dysplastic morphology in the erythroid; granulocytic or megakaryocytic lineages (at least bilineage), and (2) an acquired, sustained clonal cytogenetic abnormality with 5% or more blasts in the BM. Almost half of all children with MDS in one series presented with refractory cytopenia, most notably neutropenia and thrombocytopenia.
Morphologically, the BM may be hypocellular, normocellular, or hypercellular. A diagnosis of MDS is made based on the presence of dysplastic changes in at least two cell lineages. The dysplastic changes in the granulocytes (hypogranulation, nuclear hyposegmentation, megaloblastoid maturation, and a left shift with an increased number of myeloblasts), megakaryocytes (micromegakaryocytes, abnormal megakaryocyte nuclei), monocytes (increase in BM monocytes, abnormal granulation with persistence of azurophilic granules, hemophagocytosis, abnormal nuclei, and giant forms), or erythroid lineages (megaloblastoid maturation, nuclear budding and multinucleated forms, and ringed sideroblasts) can be multiple and varied. Similar dysplastic changes can occur in the peripheral blood for each of these lineages. Although dysplastic changes in the BM are a common feature of MDS, it is important to remember that dysplasia, unto itself, is not diagnostic of MDS because dysplastic features are associated with other conditions, and can be found in normal BM donors.
Flow cytometric analysis can serve as a useful addition to histopathology and to quantitate the number of blasts based on aberrant cell surface antigen expression and to exclude AML. It is also helpful in detecting populations of PNH-like cells. However, while beneficial, flow cytometric findings are not generally diagnostic of MDS.
Cytogenetic abnormalities are seen in approximately half of children diagnosed with de novo MDS. Karyotypic abnormalities most commonly seen are −7, 7q−, and +8. Abnormalities in chromosomes 6, 9, 11, 12, and 13 are rare in children. Deletions of chromosome 20q can be seen in pediatric patients with SDS. Specific abnormalities seen in adults, including −5, 5q−, and −Y, are very rarely seen in children.
Although the history, physical examination, evaluation of the BM and peripheral blood, and cytogenetic analysis often make the diagnosis of MDS, other diseases should be considered. Congenital disorders such as Down syndrome, FA, Shwachman-Diamond syndrome, Diamond-Blackfan anemia, congenital dyserythropoietic anemias, and hereditary sideroblastic anemia should be considered. The differential should also include AML with a low blast count, mitochondrial cytopathies such as Pearson syndrome, rheumatic diseases including juvenile idiopathic arthritis, and myeloproliferative disorders. Specifically, PNH, although rare in children, should be considered. Deficiencies of vitamin B 12 and folate can cause megaloblastic changes that resemble the dysplastic changes seen in MDS. Other nutritional deficiencies, including copper, iron, thiamine, riboflavin, and pyridoxine, should be considered. Infections caused by human immunodeficiency virus, parvovirus, Epstein-Barr virus, cytomegalovirus, and human herpes virus 6 can cause changes that resemble MDS. Finally, the differential diagnosis should include toxins (insecticides, chemotherapy agents, and arsenic), as well as cytokine exposure and radiation.
Hypoplastic MDS (RCC) can be difficult to distinguish from severe aplastic anemia and inherited BM failure (BMF) syndromes, especially when no chromosomal aberrations are detected. One interesting area of research that may help in differentiating between these diagnoses is the use of cytokine-based programs. In one preliminary study, thrombopoietin and IL-17 levels were useful in differentiating hypoplastic MDS from aplastic anemia. When the diagnosis is unclear, prospective monitoring and serial BM examinations may serve as useful aids in making an accurate diagnosis.
Although MDS is a heterogeneous, clonal disease of hematopoietic stem cells (HSCs) that can manifest different clinical courses, it is not readily curable by conventional chemotherapy, and requires allogeneic hematopoietic cell transplantation (HCT) to be cured in most cases. Some children with RCC and RCMD, who do not have life-threatening neutropenia nor require transfusions, may only require close observation (see box on Treatment Overview for Children With MDS ). Although children with this disease may eventually develop progressive disease requiring HCT, they may have long periods when minimal treatment is required. The use of AML-like chemotherapy for patients with RCMD with excess blasts (RCMD-EB) is controversial, but may serve to “debulk” patients with a high percentage of blasts before HCT. However, this potential benefit may be offset by toxicities associated with AML-like chemotherapy. Finally, patients with advanced MDS (RAEB-T or t-MDS) should be treated like those with AML.
Although a number of methods have been developed to predict the outcome of adults with MDS, the systems now most commonly used in adults are the International Prognostic Scoring System , that uses percentage BM blasts, karyotype, and number of cytopenias to assign a score that is then used to predict outcome, and the WHO classification-based prognostic scoring system (WPSS) that includes more recently identified prognostic factors (e.g., transfusion dependency and multilineage dysplasia). A recent study demonstrated that both systems “well represent” the prognostic risk for patients whose MDS is defined by the WHO classification criteria. Although these are effective tools for adults, their value for children is very limited.
Secondary MDS can develop in both children and adults after exposure to chemotherapy and radiation. Alkylating agents used to treat Hodgkin disease, non-Hodgkin lymphoma, and Ewing sarcoma are particularly concerning in children. Interestingly, there is also evidence to suggest that the cardioprotectant dexrazoxane, a topoisomerase II inhibitor with a mechanism of action that is different from etoposide and doxorubicin, may have increased the incidence of secondary MDS and AML in children treated for Hodgkin disease.
Treatment options for children with secondary MDS are limited. Although AML-like chemotherapy can induce a period of remission and reduction in BM blasts, it is not curative. As noted above, allogeneic HCT has curative potential, but outcomes remain poor, with only 20% to 30% of children surviving in the reported series. HCT has also been reported in children and young adults with secondary MDS, arising from hereditary BMF syndromes, as well as MDS secondary to chemotherapy exposure. In the setting of hereditary BMF syndromes, careful evaluation of potential familial donors needs to be factored into donor choice. HCT outcomes for MDS secondary to chemotherapy exposure have been poor, with both high rates of relapse and transplant-related mortality. Current areas of investigation include the utility of monitoring hematopoietic chimerism after transplantation, and intervening with preemptive immunotherapy to try to prevent relapse.
A multitude of agents have been studied for the treatment of myelodysplastic syndromes (MDS) in adults, but only rarely in children. These include low-dose chemotherapy (cytosine arabinoside, melphalan, hydroxyurea, etoposide, topotecan, 6-mercaptopurine, and busulfan), hormones (glucocorticoids and androgens), differentiating agents (13- cis -retinoic acid, all- trans retinoic acid), hematopoietic growth factors (granulocyte-macrophage colony-forming factor, granulocyte colony-forming factor, and erythropoietin), demethylating agents (decitabine, 5-azacytidine), proteasome inhibitors, antiangiogenic agents, and arsenic. This has resulted in three drugs (lenalidomide, azacitidine, and decitabine) that are now approved by the US Food and Drug Administration for the treatment of MDS in adults. However, as there are currently no safety or efficacy data to support the use of these agents in children with MDS, these drugs are not approved for MDS in children.
In addition, a large number of agents are being tested in clinical trials in the broad categories of kinase inhibitors, deacetylase inhibitors and DNA methyltransferase inhibitors, altered cell metabolism, cytotoxics, cell cycle inhibitors, immunomodulators and immunosuppressive agents, apoptosis modulators, and others.
The difficulty in assessing the safety and efficacy of new agents in children with MDS is illustrated by the Children’s Oncology Group’s recent prospective study of amifostine. This prospective Phase II cooperative group study was unable to be completed because of lack of accrual. As a result, the safety and efficacy of amifostine in children with MDS remains uncertain. Despite a lack of successful prospective clinical trials in children with MDS, there are limited retrospective data on the use of the hypomethylating agent azacytidine in children with newly diagnosed MDS, and in the palliative setting, children with relapsed MDS. These children had refractory cytopenia of childhood (RCC), advanced and secondary MDS, and prior to allogeneic hematopoietic cell transplantation (HCT). These retrospective reviews suggest that azacytidine is tolerable and responses are possible, although the safety and efficacy in these various clinical settings have not been demonstrated in prospective clinical trials in children.
The utility of chemotherapy prior to allogeneic HCT for childhood MDS is controversial. One retrospective report suggested that use of pre-HCT chemotherapy was associated with poorer post-HCT survival; however, interpretation was limited by the lack of randomization, such that patients with more aggressive disease may have been the ones treated with chemotherapy. Another report suggested less post-HCT relapse after pre-HCT chemotherapy in more advanced cases. Rather than using traditional cytotoxic chemotherapy, a single-center report of non-random use of azacitidine pre-HCT decreased blast counts and improved survival post-HCT. This approach has also been demonstrated to result in better outcomes after HCT for adults with MDS.
Most children with MDS require allogeneic HCT for curative therapy. Although children have been included in published HCT studies for MDS that are largely focused on adult patients, several studies have focused specifically on children. These studies suggested a probability of disease-free survival in about 50% of patients undergoing human leukocyte antigen (HLA)–matched related donor HCT. However, HCT outcomes have improved in recent years. A single-center report compared OS rates between the time-period of 1997–2007 (50%) and 2008–2017 (87%; P = .006). The authors attributed much of this improvement to advances in haploidentical donor HCT.
Given the rarity of this disease in children, the best approach for HCT is not clear. A small single center report suggested that reduced intensity conditioning (RIC) was at least equivalent to fully myeloablative conditioning (MAC) for patients with RCC. Another small single-center report also suggested high rates of EFS after RIC for patients with RCC. Treosulfan, if available, may be an especially attractive agent to use in conditioning for patients with MDS, given low toxicity and relapse rates. Conversely, patients with more advanced forms of childhood MDS, including RCC associated with monosomy 7, 7q−, or multiple genetic abnormalities, continue to have high rates of post-HCT relapse and probably should be treated with MAC. Algorithms have been proposed to help guide this decision. There are also single-arm adult trials which have attempted to prevent post-HCT relapse with the using of hypomethylating agents, though these have been fraught with high rates of toxicity, and pediatric data is lacking.
JMML is classified by the WHO as an overlap of MDS and MPN. It is an aggressive myeloid malignancy of young children with poor outcomes to conventional therapies. The diagnostic criteria are complex ( Table 65.1 ), but recent advances in elucidating the molecular genetics of the disorder demonstrate that approximately 95% of children will harbor an alteration in one of several genes in the Ras pathway. Thus, there is now international agreement on the diagnostic criteria (see Table 65.1 ), with an emphasis on incorporating these molecular criteria, as well as a recent definition of common response criteria.
Category 1 | Category 2 | Category 3 |
---|---|---|
All of the Following: | At Least 1 of the Following: | At Least 2 of the Following: |
|
|
|
a For the 7%–10% of patients without splenomegaly, the diagnostic criteria must include all other features in Category 1 AND one of the parameters in Category 2 OR no features in Category 2 but two features in Category 3.
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here